The Common European Framework in its political and educational context What is the Common European Framework?

Self- or teacher-assessment: Provided the descriptors are positive, independent statements
they can be included in checklists for self- and teacher-assessment. However, it is a
weakness of the majority of existing scales that the descriptors are often negatively
worded at lower levels and norm-referenced around the middle of the scale. They also
often make purely verbal distinctions between levels by replacing one or two words
in adjacent descriptions which then have little meaning outside the co-text of the
scale. Appendix A discusses ways of developing descriptors that avoid these problems.
2.
Performance assessment: A more obvious use for scales of descriptors on aspects of com-
petence from Chapter 5 is to offer starting points for the development of assessment
criteria. By guiding personal, non-systematic impressions into considered judge-
ments, such descriptors can help develop a shared frame of reference among the
group of assessors concerned.
There are basically three ways in which descriptors can be presented for use as assess-
ment criteria: 
•
Firstly, descriptors can be presented as a scale – often combining descriptors for dif-
ferent categories into one holistic paragraph per level. This is a very common
approach.
•
Secondly, they can be presented as a checklist, usually with one checklist per relevant
level, often with descriptors grouped under headings, i.e. under categories.
Checklists are less usual for live assessment.
•
Thirdly, they can be presented as a grid of selected categories, in effect as a set of par-
allel scales for separate categories. This approach makes it possible to give a diagnos-
tic profile. However, there are limits to the number of categories that assessors can
cope with.
There are two distinctly different ways in which one can provide a grid of sub-scales:
Proficiency Scale: by providing a profile grid defining the relevant levels for
certain categories, for example from Levels A2 to B2. Assessment is then made
directly onto those levels, possibly using further refinements like a second digit
or pluses to give greater differentiation if desired. Thus even though the perfor-
mance test was aimed at Level B1, and even if none of the learners had reached
Level B2, it would still be possible for stronger learners to be credited with B1+,
B1++ or B1.8.
Examination Rating Scale: by selecting or defining a descriptor for each relevant
category which describes the desired pass standard or norm for a particular
module or examination for that category. That descriptor is then named ‘Pass’ or
‘3’ and the scale is norm-referenced around that standard (a very weak perfor-
mance = ‘1’, an excellent performance = ‘5’). The formulation of ‘1’ & ‘5’ might be
Assessment
181

---

other descriptors drawn or adapted from the adjacent levels on the scale from
the appropriate section of Chapter 5, or the descriptor may be formulated in rela-
tion to the wording of the descriptor defined as ‘3’.
9.2.3
Describing the levels of proficiency in tests and examinations to aid comparison
The scales for the Common References Levels are intended to facilitate the description of
the level of proficiency attained in existing qualifications – and so aid comparison
between systems. The measurement literature recognises five classic ways of linking sep-
arate assessments: (1) equating; (2) calibrating; (3) statistical moderation; (4) benchmark-
ing, and (5) social moderation.
The first three methods are traditional: (1) producing alternative versions of the same
test (equating), (2) linking the results from different tests to a common scale (calibrating),
and (3) correcting for the difficulty of test papers or the severity of examiners (statistical
moderation). 
The last two methods involve building up a common understanding through discus-
sion (social moderation) and the comparison of work samples in relation to standardised
definitions and examples (benchmarking). Supporting this process of building a common
understanding is one of the aims of the Framework. This is the reason why the scales of
descriptors to be used for this purpose have been standardised with a rigorous develop-
ment methodology. In education this approach is increasingly described as standards-
oriented assessment. It is generally acknowledged that the development of a
standards-oriented approach takes time, as partners acquire a feel for the meaning of the
standards through the process of exemplification and exchange of opinions.
It can be argued that this approach is potentially the strongest method of linking
because it involves the development and validation of a common view of the construct.
The fundamental reason why it is difficult to link language assessments, despite the sta-
tistical wizardry of traditional techniques, is that the assessments generally test radically
different things even when they are intending to cover the same domains. This is partly
due to (a) under-conceptualisation and under-operationalisation of the construct, and